A high memory usage alert would be useful please.  I've run into a situation a few times where things have gone bad because I've run out of memory - I can see it coming now and restarting a few dockers to release memory helps, but an alert would be nice for when I'm not monitoring.

This seems like something that could be done thru a plugin

I'm not picking on you specifically, but I absolutely hate this answer.

 

Something as basic as an out of memory warning should absolutely be handled by a server OS, and should 100% not require a community member to implement.

unRAID has a very small development team, so it's completely fine that they haven't got around to adding this yet, but that doesn't mean the userbase should get comfortable with the "there's a plugin/script for that" mentality.

 

 

Yea, you are right, specific "standard" options should be included via OS not via plugins.

But that is how it is.

On 5/2/2018 at 8:32 AM, nuhll said:

Yea, you are right, specific "standard" options should be included via OS not via plugins.

But that is how it is.

And how it was the whole history of unraid. We didn't even have email notifications without 3rd party plugins for many years.

 

If a specific feature is asked for often enough, and is possible to implement, it will get done, eventually. You may have to wait a few months or years, but it will eventually happen. Look on the bright side, limetech has yet to ask for an upgrade fee for existing license holders to access new features and functionality, unlike almost EVERY other software company around.

 

I'll take slow incremental software changes, reliability and stability over fast changes and upgrade fees any day of the week.

 

You ask for all these changes, and seem to take offence when people don't immediately agree with you that your specific needs should direct the focus of the company. That seems very selfish and shortsighted to me. Unraid has many different uses and environments, and any changes need to take into account a whole range of customers, not just your specific use case.

Just to put stuff into perspective, the high memory usage situation mainly reared its head with the inclusion of dockers and VMs.

Even before with just plugins, people weren't running all that much and server applications usually made do with whatever ram was available.

But with containers and VMs and people running desktops and stuff, people sometimes forget to spec out the memory requirements properly and the whole thing falls apart due to the absence of swap.

 

while it just came a thing after dockers and so on were implemented... a "user" uses RAM scenario should never be able to freeze the system. (look  at windows, last startet app crash with too low RAM)

Edited May 9, 20188 yr by nuhll

5 hours ago, nuhll said:

while it just came a thing after dockers and so on were implemented... a "user" uses RAM scenario should never be able to freeze the system. (look  at windows, last startet app crash with too low RAM)

 

When the end users are allowed to configure how much RAM to allocate to different additional features, it's impossible for LT to make a system where the users can't run out of RAM.

 

And with the root FS in RAM, the only way to stop people from running out of RAM by writing to the root FS is to limit the size of the root FS - but while that stops people from consuming all RAM, it doesn't solve the issue with a full root FS and all the problems that leads to.

 

In the end, it's impossible to make an idiot proof system - nothing is foolproof to a sufficiently talented fool.

Users running out of ram was very much a thing on the 32bit unraid versions, years before dockers and vms on unraid were a thing.

@limetech any plans for something like this to be implemented?

Hey all,

great thread, valid points on Unraids support, but yeah, i'll take a one time payment and limited actual official updates. 

 

what i have done is used home assistant to send the notifications, I've had memory leaks before, still trying to debug what actually caused it.

essentially i use glances docker app and then push the data to home assistant, when the memory on unraid reaches 90% i get it to send me a notification saying my undraid server has reached 90% memory usage so i know i need to look into it before it crashes unraid.

 

let me know if you want some more detail on how.

 

here is a screenshot in home assistant with the memory issue historical shown.

 

 

 

Does anyone know how to create a notification of memory usage at x%? I am trying to debug my server crashing and i haven't gotten much helpful responses so far across unraid forum / discord servers besides "turn off the app and see if it works". I unfortunately need to move apartments in 1 week so i need to desperately get this crash-free before i leave. I want to diagnose this and have an alert setup that when memory reaches 70%, 80%, 90%, i can quickly login and see via HTOP what is leaking.

Edited January 17, 20242 yr by Linguafoeda

It would be possible write a script that checks memory usage using the linux 'free' command and set this to run regularly with the User Scripts plugin.  The 'ps' command might also be relevant if you want to monitor particular processes.  What you do when you detect a trigger level will be up to the script - although it could either write entries to the syslog, or alternatively raise a notification at the Unraid level.

4 hours ago, itimpi said:

It would be possible write a script that checks memory usage using the linux 'free' command and set this to run regularly with the User Scripts plugin.  The 'ps' command might also be relevant if you want to monitor particular processes.  What you do when you detect a trigger level will be up to the script - although it could either write entries to the syslog, or alternatively raise a notification at the Unraid level.

 

I would love to get an unraid notification (which then feeds to my discord). Do you have any idea how to setup the script?

I will try to setup discord memory usage alerts with netdata container as reported here: